Enterprise Productivity

Madhavi Rajan shared how Rackspace uses AI for sales augmentation, churn prediction, support workflows, and HR knowledge bots, while navigating employee concerns and compliance requirements across industries like healthcare.

One comment captured the near‑future vision:

“We’re close to agents doing overnight work and reporting back in the morning, governance just needs to catch up.” – David Hefter, BlackRock

The panel agreed that the next phase of AI adoption will be shaped by cost discipline and architectural efficiency.

Madhavi Rajan noted that organizations are beginning to measure:

• Cost per task
• Token efficiency
• Model routing strategies
• Task‑tuned model performance
• GPU utilization

Instead of simply experimenting with the latest model.

Another panelist summarized it well:

“People will start measuring themselves not by what they built, but by how efficiently they run it.”

The AI reset places a premium on sustainable architecture, not runaway experimentation.

The Emerging Battle: System‑of‑Record vs. System‑of‑Engagement

A forward‑looking discussion led by Kyle Hutchinson explored how AI agents may change where work happens inside organizations.

Traditionally:

• CRMs and ERPs = systems‑of‑record
• Email/productivity tools = systems‑of‑engagement

Now, as agents write emails, update records, run workflows, and orchestrate tasks across tools, AI is becoming the engagement layer, with CRMs acting increasingly as back‑end databases rather than primary user interfaces.

“If agents handle the work across systems, the CRM becomes the database, not the workspace.” –  Kyle Hutchinson

This shift has broad implications for pricing models, software positioning, and workflow design across the deal ecosystem.

The Mid‑Market Opportunity: Smart Growth, Not Blind Adoption

While enterprises blaze ahead, mid‑market companies may hold the greatest upside, precisely because many have yet to modernize.

Rather than trying to leap to full AI maturity, the panel recommended anchoring efforts around:

• Private, governed data
• High‑ROI, workflow‑specific use cases
• Scaffolded agentic steps, not free‑form chat
• Human‑in‑the‑loop controls where accuracy matters most
• Measurable unit economics

Matt Keep emphasized that grounded, domain‑specific AI, not generic prompts, is what moves the needle in dealmaking.

The Takeaway: The AI Reset Is a Good Thing

Across the discussion, the theme was unmistakable: the AI reset isn’t a slowdown, it’s a maturation.

Organizations are moving from:

• demos → workflows
• hype → integration
• generic models → domain‑tuned agents
• experimentation → measurement

And that is exactly where durable enterprise value gets created.

“Real value comes from data readiness, workflow design, and proven outcomes, not just the promise of models.” – Matt Keep